You can not select more than 25 topics
Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
README.md 1.2KB
README.md 1.2KB
Fake News Detection Using LLMs
📌 Description
This repository contains the implementation of the paper “Toward Fair and Effective Fake News Detection: Assessing Large Language Models.” The project focuses on evaluating the fairness and efficiency of Large Language Models (LLMs) in detecting fake news using a dataset of news articles classified by political leaning.
🚀 Features
- Analysis of LLM biases in news classification
- Evaluation of model fairness and accuracy
- Benchmarking multiple LLMs (GPT-4o, LLaMa, Qwen, Deepseek)
- News leaning classification (Democrat, Republican, Neutral, Varies)
- Fake news detection using a labeled dataset
📖 Usage
use the news_dataset.csv and news_leaning_dataset.csv as dataset and log_processor.py as processor for the outputs of each LLM.
🛠️ Technologies Used
- Python
- open-ai
- Scikit-learn
- Pandas & NumPy
📊 Dataset
The dataset consists of news articles labeled with political leanings and fact-checking results. The files include:
- news_dataset.csv: Contains raw news articles with metadata with labeled POV.
- news_leaning_dataset.csv: Labels news articles as Democrat, Republican, Neutral, or Varies with Labeled leanings.